Picture a poodle that spent its childhood on all fours, then hit adolescence and decided the front legs were optional.

Something close to that appears to have happened in what is now northern Arizona, about 215 million years ago, to a beaked reptile sitting on the crocodile side of the family tree. Elliott Armour Smith and Christian Sidor of the University of Washington have named it Sonselasuchus cedrus, and their description of it appears in the Journal of Vertebrate Paleontology.

What came out of the bonebed

Nobody dug up a tidy skeleton. Since 2014, teams working in Petrified Forest National Park have pulled more than 3,000 fossil bones out of a single deposit, and over 950 of them belong to this one animal. Sidor, who has led the fieldwork there for more than a decade, told the Taylor & Francis newsroom that the deposit keeps turning up new material, among it fish, amphibians, dinosaurs and other reptiles found in the same rock.

Reconstructing an animal from a jumble of loose bones is a particular kind of headache, because there is no guarantee that any two of them came from the same individual. What you get instead is a population: lots of femurs, lots of humeri, at a range of sizes.

That, as it turns out, is exactly the sort of pile you want if the question is how an animal changed shape as it aged.

The tell was in the leg bones

Armour Smith and Sidor compared limb proportions across those different-sized individuals and found the ratio shifting with size. Small animals had forelimbs and hindlimbs that were roughly in balance. Larger ones had hindlimbs that were longer and heavier relative to everything else.

Biologists have a name for growth that runs at different rates in different body parts: allometry, and it is common enough across living animals. What is unusual here is where it appears to lead. As Armour Smith explained in the research statement carried by Phys.org, the creatures likely started life on four legs and moved onto two as they grew up.

Precision matters here.

Nobody watched this happen. It is an inference from bone proportions in a disarticulated assemblage, made by two researchers in one paper, and it will stand or fall on whether other shuvosaurid material shows the same pattern.

This sort of shift turns out to be fairly common

Humans do this, which is the easiest example going: a crawling toddler and a walking five-year-old are the same animal with different proportions and a different centre of mass. Spiny-tailed iguanas do a version of it minute to minute, rising onto their hind legs to hit top speed. And in 2019 a team analysing the sauropodomorph dinosaur Mussaurus patagonicus reported in Scientific Reports that its centre of mass migrated backwards towards the pelvis as it grew from a 60-gram hatchling to a tonne-plus adult, shifting how it stood.

So the mechanism itself is unremarkable. What makes Sonselasuchus worth a second look is the branch it turns up on, a lineage where nothing alive does anything of the sort today.

Crocodile relatives were never the dull ones

“Living fossil” gets thrown at crocodiles constantly and it flattens 200 million years of experimentation into a single sleepy image. Sonselasuchus belonged to the shuvosaurids, a small group within Pseudosuchia, the broader crocodile-line branch of the archosaur family tree whose only living members today are crocodiles and alligators. Sci.News reports it stood around 63.5 centimetres tall, with a toothless beak, an oversized eye socket and hollow bones.

If that sounds like an ostrich-mimic dinosaur, that is the point. Sterling Nesbitt and Mark Norell made the case for extreme convergence back in 2006, describing a close relative called Effigia okeeffeae in Proceedings of the Royal Society B. Features of its skull, pelvis, tail and vertebrae matched ornithomimid theropods almost point for point, and the two lineages had split some 250 million years earlier. Croc-line and bird-line animals were sharing ecosystems, competing for the same jobs, and arriving independently at the same body plan.

The genus name, as Smithsonian magazine notes, comes from the Sonsela Member, the rock unit it was found in. The species name comes from the cedar, evoking the conifer forests where it lived.

Ten years of digging, 3,000-odd bones, and the quarry is still handing over animals that did not have names. Whatever surfaces next probably moves in a way nobody has thought to predict either.